Experimental Hematology 27 (1999) 353–364

0301-472X/99 $–see front matter. Copyright © 1999 International Society for Experimental Hematology. Published by Elsevier Science Inc.PII S0301-472X(98)00038-1

SIGNAL TRANSDUCTION

Differential effect on U937 cell differentiation by targeting transcriptional factors implicated in tissue- or stage-specific induced integrin expression

Angel García, Alfonso Serrano, Eduardo Abril, Pilar Jimenez,Luis Miguel Real, Julia Cantón, Federico Garrido, and Francisco Ruiz-Cabello

Departamento de Análisis Clínicos, Hospital Universitario Virgen de las Nieves, Universidad de Granada, Granada, Spain

(Received 24 February 1998; revised 19 June 1998; accepted 6 July 1998)

The inhibition of transcription factor functions was used to de-fine their role in phorbol ester-induced cellular differentiationof a monocytic cell line, U937. We demonstrate a differentialeffect on cell adhesion and differentiation: antisense or com-petitive binding with double-stranded oligonucleotides antago-nized the functions of AP-1, NF-

k

B, and PU.1 transcriptionalfactors. In the presence of phorbol 12-myristate 13-acetate(PMA), U937 cells attached to the plastic surface and cells werecharacterized by marked expression of

b

2-integrin moleculeson the cell surface. We show that the in vivo differentiation ofU937 cells appears to occur normally in the absence of AP-1activity. In contrast, the addition to the cell culture of phospho-rothioate oligonucleotides that contained the NF-

k

B or PU.1binding sites significantly inhibited U937 differentiation. Theabsence of NF-

k

B led to pleiotropic effects with a clear reduc-tion in the expression of integrin and other lineage-specific my-eloid antigens on the cell surface. In contrast, the absence ofPU.1 had a more restricted effect on integrin expresion on thecell surface, probably as a result of blockage of CD18 geneexpression. © 1999 International Society for ExperimentalHematology. Published by Elsevier Science Inc.

Keywords:

Antisense oligonucleotide—Cell differentiation—

Integrin—Cell adhesion

Introduction

U937, a histiocytic lymphoma cell line with monoblasticcharacteristics, can be induced to differentiate towards mor-phologically mature macrophage-like cells after treatmentwith phorbol esters [1–3]. Phorbol esters such as phorbol-12-myristate 13-acetate (PMA) have pleiotropic effects oncells [4]. Phorbol exerts its biologic effects by altering geneexpression through the activation of PKC [5,6] and modu-

lating the activity of transcriptional factors that bind cis ele-ments, such as NF-

k

B [7–9], AP-1 [10,11], AP-2 [12,13],AP-3 [14], and Ets [15]. In this process, the expression ofadhesion receptors changes notably: members of the

b

2-integrin family increase during differentiation in U937 andHL-60 cells [2,6]. The CD11/CD18 (

b

2-integrin) familyconsists of three heterodimeric cell surface receptors(CD11a/CD18, CD11b/CD18, and CD11c/CD18) that showrestricted expression on leukocytes [16]. Members of thisfamily are activated by PMA and a variety of other physio-logic inflammatory stimuli, and play a critical role in leuko-cyte-endothelium interactions during immune and inflam-matory reactions. These molecules are also responsible forthe establishment of firm adhesion and transendothelial mi-gration [17].

The level of expression of

b

2-integrins is under criticalcontrol during myeloid differentiation and inflammation.Some elements of this control are exerted at the transcrip-tional level, and a coordinated transcriptional up-regulationof the integrin subunits CD11/CD18 take places. Impair-ment of the expression of CD18 in an inherited immunode-ficiency (Leu-CAM deficiency) alters the stoichiometry andabolishes cell surface expression [17].

The molecular events that ultimately influence the tissue-and differentiation-specific transcription of these genes dur-ing macrophage differentiation are not totally clear. Thegenes encoding both CD18 and CD11 subunits share simi-larities in the organization of their promoters and in the na-ture of the transcriptional factors, supporting a role for someof them in lineage commitment and coordinated expressionduring differentiation [15]. In this context, comparisons ofthe integrin promoters for CD11a, CD11b, CD11c, andCD18 reveals the presence of AP-1 and Ets-binding motifs[15,18–24]. These nuclear factors may be responsible forcoordinated expression of these proteins during myeloiddifferentiation. Induction of these cells with PMA increasesthe transcriptional activity of AP-1 [25], NF-

k

B [8–11], andPU.1 [26]. In the present study, we used double-stranded (ds)

Offprint requests to: Francisco Ruiz-Cabello, PhD, Servicio de AnálisisClínicos e Inmunología, Hospital Universitario Virgen de las Nieves, Uni-versidad de Granada, Avda. Constitución s/n, 18014 Granada, Spain; E-mail:fruizc@goliat.ugr.es

354

A. García et al./Experimental Hematology 27 (1999) 353–364

phosphorothioate oligonucleotides that containied AP-1, Ets,or

k

B consensus sequences to inhibit sequence-specific DNA-binding proteins during differentiation in U937 cells as amodel of monocytic maturation. We show that whereas U937cells may be induced to differentiate in the absence of ap-parent AP-1 activity, the sequestration of Ets factors andNF-

k

B has a profound effect on cell attachment and the ex-pression of

b

2-integrin molecules.

Materials and methods

Cell culture, reagents, and induction

U937 monoblastic cells were obtained from the ATCC (Rockville,MD). Cells were cultured in RPMI 1640 (Gibco, Paisley, UK) sup-plemented with 5% FCS, 2 mM glutamine, and 50

m

g/mL gentam-icin at 37

8

C in a humidified atmosphere with 5% CO

2

. To inducedifferentiation, U937 cells were resuspended in fresh medium (4

3

10

5

/mL), cultured for 12 hours, and then treated with PMA (20 ng/mL) (Sigma, St. Louis, MO) for up to 5 days. BisindolylmaleimideGF109203X [27] (Boehringer Mannheim, Barcelona, Spain) wasused at a concentration of 5

m

M, curcumin [28] (Sigma) at a con-centration of 2.5–5

m

M, and staurosporine [29,30] (Sigma) at aconcentration of 50 nM. The cells were cultured for 0.5–1.5 hoursbefore induction with PMA. Every 24 hours the culture mediumwas changed and a new dose of inhibitors was added. At the con-centrations used, the inhibitors did not affect cell viability.

Phosphorothioate oligodeoxynucleotides

All oligodeoxynucleotides used in antisense and decoy studieswere phosphorothioate, and were synthesized in solid phase ac-cording to the modified phosphoramidate method on a Beckmanmodel Oligo 1000 automated DNA synthetizer (Beckman Instru-ments Inc., Palo Alto, CA). For stepwise thiathion phosphite link-ages, the oxidation step was in 0.2 M [3H]-1,2-benzodithiole-3-one1,1-dioxide (Sigma) in acetonitrile [31]. Oligonucleotides were an-alyzed by polyacrylamide gel electrophoreis on denaturing gelsand material judged to be at least

.

90% full-length was used. Oli-gonucleotides were depyrogenated by ultrafiltration. Sequences ofthe antisense and sense oligonucleotides used on this study are de-picted in Table 1. Oligonucleotide concentrations were calculatedassuming 1 OD

260

unit

5

33

m

g/mL. Phosphorothioate oligonucle-otides for competitive binding experiments (decoy experiments)were obtained in the same way. Double-stranded oligonucleotideswere formed at a final concentration of 10 mM in TE pH8 by heat-ing at 95

8

C for 3 minutes and cooling slowly during 90 minutes toroom temperature.

Cell transfections and incubations with oligonucleotides

In transfections with liposomes, the cells were grown in T-75flasks until 1

3

10

6

cel/mL. At this time, 5 mL of cells werewashed in serum-free Opti-MEM medium (Gibco) prewarmed to37

8

C. The cells were then resuspended in Opti-MEM containing 10

m

g/mL DOTMA/DOPE solution (Lipofectin) (Gibco) and the ap-propriate sterilized oligonucleotide at a concentration of 10

m

M.The cells were incubated for 4 hours at 37

8

C, and the medium wasremoved and replaced with normal growth medium containing 20ng/mL PMA and a fresh dose of oligonucleotide. Every 24 hoursthe culture medium was changed and a new dose of oligonucle-otide was added. Both basal and PMA-treated cells were pretreated

with DOTMA/DOPE for 4 hours at 37

8

C in the absence of oligo-nucleotides.

Transfections by electroporation were performed with a BTX600electroporator, as previously described [32]. U937 cells (1 to 2

3

10

7

) were electroporated (100 V, 2950

m

F, 186 Ohms) in 500

m

LRPMI containing 10

m

M of appropriate oligonucleotide. Bothbasal and PMA-treated cells were electroporated in the absence ofoligonucleotide. Every 24 hours the culture medium was changedand a new dose of oligonucleotide was added.

Incubations of cells with antisense or sense oligonucleotideswere performed by adding the corresponding oligonucleotide tothe culture medium 1-4 hours before induction with PMA. In de-coy experiments, the double-stranded phosphorothioate oligonu-cleotides were added to the cells 2–4 hours before inductionwith PMA. Every 24 hours the culture medium was changed anda new dose of oligonucleotide was added. At the concentrationsused, the oligonucleotides did not affect cell viability.

Immunofluorescence analysis

After induction with PMA for 48–72 hours, cells were har-vested, washed twice in ice-cold phosphate-buffered saline(PBS), and incubated at 4

8

C with the following antibodiesmarked with fluorescein isothiocyanate (FITC) or phycoerythrin(PE): anti-CD11b-PE, anti-CD11c-PE, anti-CD15-FITC, anti-CD33-PE, anti-CD14-PE, anti-CD11a-FITC, and anti-CD18-FITC. All were purchased from Becton Dickinson (San Jose,CA). As a negative control, we used an irrelevant PE- or FITC-conjugated monoclonal antibody. Cells were analyzed by flowcytometry using a Facsort cytometer (Beckton Dickinson). Aminimun of 10

3

10

3

cells were analyzed in each experiment.

Table 1.

Oligonucleotides used in this study

Oligonucleotide Sequence

AP-1 5

9

GTGACTCATGACTCATGACTCATGACTC 3

9

3

9

ACTGAGTACTGAGTACTGAGTACTGAGT 5

9

NF-

k

B 5

9

GCATTAGGGGGCTTCCACGGCCTGA 3

9

3

9

GTAATCCCCCGAAGGTGCCGGACTT 5

9

SP-1 5

9

GGTCAGGACCCGCCCCTTCTGGTCCG 3

9

3

9

CAGTCCTGGGCGGGGAAGACCAGGCT 5

9

CRE 5

9

GGTGATTGCCTGACGTCAGAGAGCAT 3

9

3

9

CACTAACGGACTGCAGTCTCTCGTAT 5

9

antisense c-fos

1

Sense c-fos

1

5

9

GCCCGAGAACATCATCGT 3

9

5

9

ACGATGATGTTCTCGGGC 3

9

antisense c-jun

1

sense c-jun

1

5

9

CGTTTCCATCTTTGCAGT 3

9

5

9

ACTGCAAAGATGGAAACG 3

9

AP-1ppt

1

5

9

GCCCCCTCTGACTCATGCTGACA 3

9

3

9

GGGGGAGACTGAGTACGACTGTT 5

9

NF-

k

B-ppt

1

5

9

GTACATTAGGGGGGCTTCCACGGCCTG 3

9

3

9

ATGTAATCCCCCCGAAGGTGCCGGACT 5

9

PU.1-ppt

1

5

9

GCTGAAAGAAGGGCAGAAAAGGAGAAGTAG 3

9

3

9

GACTTTCTTCCCGTCTTTTCCTCTTCATCT 5

9

AP-1ppt-mut

1

5

9

GCCCCCTCTAGCAGATGCTGACA 3

9

3

9

GGGGGAGATCGTCTACGACTGTT 5

9

NF-

k

B-ppt-mut

1

5

9

GTACATTAGCTATGAGAACACGGCCTG 3

9

3

9

ATGTAATCGATACTCTTGTGCCGGACT 5

9

PU.1-ppt-mut

1

5

9

GCTGAAAGAAGGGCAGCCCGGGAGAAGTAG 3

9

3

9

GACTTTCTTCCCGTCGGGCCCTCTTCATCT 5

9

1

Phosphorothioate oligonucleotides.

A. García et al./Experimental Hematology 27 (1999) 353–364

355

Both basal and PMA-treated cells were pretreated with DOTMA/DOPE for 4 hours at 37

8

C or electroporated in the absence of oli-gonucleotides.

Immunoprecipitation analysis

Cells (5

3

10

6

) were incubated for 1 hour in 1 mL [

35

S] methion-ine-free RPMI (Gibco) supplemented with 10% dialysed fetal calfserum, and [

35

S]-methionine (250

m

Ci) (Amersham) with antisenseor sense oligonucleotides and then stimulated with PMA (20 ng/mL) during 1 hour. Cells were washed in PBS and lysed in 10 mMTris-HCl, pH 7.4, 150 mM NaCl, 1 mM EDTA, 1% Triton X-100,1 mM phenylmethyl sulfonyl fluoride (PMSF), and 1% bovine se-rum albumin. After lysis, supernatants were cleared by ultracentrif-ugation at 10,000

g

for 1 hour. Lysates were then adjusted forequal incorporation of [

35

S]-methionine in trichloroacetic acid-pre-cipitable counts and processed by immunoprecipitation. Lysateswere precleared with 3

m

l normal mouse serum and then with 100

m

l protein-A-sepharose 10% (w/v) (Pharmacia, Uppsala, Sweden).Precleared lysates were incubated with 10

m

g of a monoconal anti-body anti-human c-FOS, Ab-1, (Oncogene Science Inc., NY)

,

overnight at 4

8

C. Immune complexes were isolated with protein-A

Sepharose for 1 hour on ice, then washed in lysis buffer. Pelletswere resuspended in 40

m

L SDS sample buffer and loaded on a12% SDS-PAGE gel under reducing conditions.

Electrophoretic mobility shift assay

Cells lines were grown to log phase before harvest for nuclear pro-tein extraction. Nuclear proteins were extracted 6–24 hours after in-duction with PMA and were prepared as previously described [32]with modifications [33]. Briefly, 10

3

10

6

cells were collected,washed twice in PBS, and resuspended in 400

m

L cold buffer A (10mM HEPES, pH 7.6, 10 mM KCl, 1 mM dithiothreitol, 0.5 mM phe-nylmethylsulfonyl fluoride, 0.5 mM spermidine, 0.15 mM spermine)with 25

m

L NP-40 10% and allowed to swell for 15 minutes on ice.The nuclei were collected by centrifugation at 400

3

g

for 5 min-utes, washed once in buffer A, and resuspended in 50

m

L of buffer C(20 mM HEPES, pH 7.9, 0.4M NaCl, 1 mM EDTA, 1 mM EGTA, 1mM dithiothreitol, 1 mM phenylmethylsulfonyl fluoride). All buff-ers contained aprotinin (2

m

g/mL), leupeptin (2

m

g/mL), and pepsta-tin (1

m

g/mL) as additional protease inhibitors. The tubes were vig-orously rocked on a shaking platform at 48C for 30 minutes andcentrifuged at 12,000 3 g for 5 minutes at 48C, and the supernatant

Figure 1. Effect of different treatments on PMA-induced differentiation of U937 cells. Cells were pretreated with each agent, bisindolylmaleimideGF109203X (5 mM), staurosporine (50 nM), phosphorothioate antisense c-fos oligonucleotide (10 mM), phosphorothioate antisense c-jun oligonucleotide (10mM), double-stranded phosphorothioate AP-1 oligonucleotide (7.5 mM), double-stranded phosphorothioate NF-kB oligonucleotide, or double-stranded phos-phorothiate PU.1 oligonucleotide, and then stimulated with PMA (20 ng/mL). The effects on morphologic changes and homotypic adhesion are compared.Phase-contrast photomicrophographs were made after 72 hours of PMA induction. Results from one of three experiments.

356 A. García et al./Experimental Hematology 27 (1999) 353–364

was aliquoted and rapidly frozen on liquid N2 and stored at 2808C.The protein concentration of the extracts was determined with theBradford protein assay (Bio-Rad, Madrid, Spain).

Binding activity was assessed using DNA oligomers, some ofthem corresponding to the promoter region of b2-integrins (Table1). Complementary oligodeoxynucleotides containing the requiredbinding sequence were synthesized in our laboratory as describedabove. Probes were annealed with complementary strand by heat-ing to 958C for 3 minutes, then slowly cooling to room temperature.

One-hundred-nanogram amounts of double-stranded oligonu-cleotides were labelled by filling-in with Klenow fragments and

[a-32P] dCTP or dATP (3000 Ci/mmol). A quantity amounting to0.5 ng (3 3 104 cpm) of labeled oligomers was then incubated at48C for 20 minutes with 2–6 mg of nuclear extract in 20 mL ofbinding buffer containing 5 mM HEPES (pH 7.9), 6 mM MgCl2,

0.2 mM EDTA, 15 mM KCl, 15% glycerol, 1 mM DTT, and 2 mgpoly(dI-dC). In certain experiments, supershift analysis was per-formed using c-fos- and jun-specific antisera, kindly provided byDr. Rodrigo Bravo (Bristol-Myers-Squibb Research Institute, Prin-ceton, NJ). For these experiments, the cellular extract was preincu-bated for 30 minutes at 48C with 0.5 mL of antisera before the la-beled probe was added. In competition experiments, a 100-fold

Figure 2. (A) Expression of cell surface antigens in U937 cells after differentiation with PMA. Immunofluorecence flow cytometry was performed onuntreated cells (Basal) or cells induced to differentiate with PMA (20 ng/mL) during 72 hours. Cells were washed and stained with either FITC- or PE-labelledmAbs against the following surface markers: CD11a, CD11b, CD11c, CD18, CD14, CD15, and CD33. Stained cells were analyzed on a FACSort cytometer.Black areas show nonstimulated cells and white areas show staining after PMA induction. (B) Induction by PMA of transcriptional factor binding activity inU937 cells. Radiolabeled oligonucleotides AP-1, NF-kB, SP-1, and CRE (Table 1) were incubated with nuclear extract (3 mg) prepared from U937 cells(Basal lanes), or U937 cells induced to differentiate along the monocytic pathway with PMA during 24 hours (PMA lanes). The last binding reactions wereperformed in the presence of a 100-fold excess of the appropriate unlabeled oligonucleotide (Comp lanes).

A. García et al./Experimental Hematology 27 (1999) 353–364 357

excess of unlabeled oligomer was incubated at 48C with the cellu-lar extract for 15 minutes before the radioactive probe was added.An oligonucleotide containing the oct-1 binding site was used tocontrol the quality of nuclear extracts (Table 1). Reactions wereelectrophoresed through 4%–6% low-ionic-strength acrylamidegels (29/1), 0.5 3 TBE (1 3 TBE is 89 mM Tris-base, 89 mM bo-ric acid, and 2mM EDTA), fixed in 10% acetic acid, dried, and ex-posed on XAR film (Kodak) at 2808C.

Results

Expression of CD11/CD18 molecules during U937 cell differentiationAnalysis of adhesion, attachment to plates and integrin ex-pression was used to evaluate differentiation of U937 cells.In the absence of PMA, nondifferentiated U937 cells did notattach to the plastic surface, and remained in an undifferen-tiated state without homotypic adhesion (Fig. 1). PMA in-duced rapid adhesion to the plastic surface accompanied bya significant increase in cell surface expression of CD11/CD18 molecules and the myelomonocytic antigen CD14(Fig. 2A). In contrast, CD33 expression was not modifiedand CD15 expression was clearly down-modulated. Theseantigens were studied to rule out any effect on cell viabilityduring oligonucleotide treatment. The increase in CD11/CD18 expression in response to PMA correlated withchanges in binding activity of nuclear transcriptinal factors.AP-1 and NF-kB binding activity was strongly increased inPMA-treated cells, whereas binding for CRE and SP-1 fac-tors was not (Fig. 2B).

Inhibition of PKC and changes in adhesion properties and integrin expressionTwo PKC inhibitors, staurosporine [29,30] and bisindoly-maleimide GF109203X [27], were investigated for theirability to inhibit U937 differentiation. Cells were incubatedfor 60 minutes in the presence or absence of 50 nM of stau-rosporine and 5 mM of bisindolylmaleimide, respectively,and then with 20 ng/mL PMA. Cellular fractions were sepa-rated for immunofluorescence studies with monoclonal an-tibodies, and CD11/CD18 expression was analyzed after 72hours.

Figure 1 shows the effect of PKC inhibitors on the mor-phology and adhesion of U937 cells. The differentiated phe-notype of U937 cells after PMA treatment was completelyblocked in the presence of staurosporine or bisindolylmale-imide, and the effect on cell morphology and homotypic ad-hesion correlated with the absence of changes in CD11/CD18 expression (Fig. 3A).

The effect of PKC inhibitors on NF-kB and AP-1 bind-ing activities was also studied. Figure 3B shows that selec-tive blockage of PKC-dependent pathways abrogated NF-kBand AP-1 PMA-induced activities: PMA induced NF-kBand AP-1 activity in electrophoretic mobility shift assays

Figure 3. (A) Changes in the expression of cell surface differentiationantigens in PMA-treated U937 cells treated with PKC inhibitors or cur-cumin. U937 cells were pretreated with bisindolylmaleimide GF109203X(Bis-GF) (5 mM), staurosporine (50 nM), or curcumin (5 mM) 1 hourbefore PMA induction. The inhibitors were added every 24 hours thereaf-ter. Seventy-two hours after PMA induction, the cells were washed andstained with FITC- or PE-mAbs. Stained cells were analyzed on a FAC-Scan cytometer. The order of the curves is the same as for CD11a. (B)Effects of PKC inhibitors and curcumin on AP-1 and NF-kB binding activ-ity in PMA-treated U937 cells. Radiolabeled oligonucleotides AP-1 andNF-kB (Table 1) were incubated with nuclear extract (3–5 mg) preparedfrom U937 cells (Basal lanes), PMA-treated U937 cells in the absence ofunlabeled specific competitor DNA (lanes PMA 1 and 2; 1 and 20 ng ofPMA, respectively) or in the presence of a 100-fold excess of unlabeledspecific competitor (comp lanes), bisindolylmaleimide GF109203X-pre-treated U937 cells treated with 20 ng/mL PMA (Bis lanes), curcumin-pre-treated U937 cells with 20ng/mL PMA (lanes curcumin 1 and curcumin 2;0.5 mM and 5 mM curcumin, respectively), or staurosporine-pretreatedU937 cells treated with 20 ng/mL PMA (Staurosporine lanes). The inhibi-tors were added 1 hour before induction with PMA, and nuclear extractswere obtained 6–24 hours after induction with PMA. Nuclear extracts werealso incubated with radiolabeled oligonucleotide oct-1 (Table 1), to verifythat the amount of extract used in each reaction was the same. Results fromone of three experiments.

358 A. García et al./Experimental Hematology 27 (1999) 353–364

(EMSA), and the binding of both factors was strongly re-duced in PKC inhibitor-treated cells. We next studied thecontribution of AP-1 and NF-kB binding activities to changesin adhesion properties and b2-integrin expression. Ets bind-ing activity was investigated because other studies of theregulatory region of CD11c and CD18 suggested that thiselement is involved in PMA-induced macrophage differen-tiation [13,26,34].

AP-1 binding activity is not required for PMA-induced U937 differentiationPrevious studies have shown that PMA treatment of U937cells is associated with c-fos and c-jun induction [35,36].AP-1 sites have been identified within CD11c and CD18promoters. This nuclear factor may contribute to the up-reg-ulation of surface-coordinated expression of these subunits.We used different oligonucleotide strategies to block AP-1activity and monitored b2-integrin expression and U937cell adhesion after either antisense or double-strand se-quence-specific phosphorothioate oligonucleotides treatment.Two antisense-specific oligonucleotides for c-fos and c-junwere tested and the two corresponding sense oligonucle-otides were used as controls (Table 1). Exposure to the junand fos antisense oligonucleotides during 4 hours before theaddition of PMA did not significantly reduce CD11/CD18expression (not shown). In fact, differentiation of U937 cellsappeared to occur normally, with morphologic features in-dicative of macrophage differentiation at 72 hours post-PMA treatment (Fig. 1). This suggested that c-jun and andc-fos induction was not essential for U937 cell differentia-tion. We then examined the expression of c-jun and c-fosmRNA (not shown), c-fos protein, and AP-1 activity (Fig.4). We found that c-fos protein expression was inhibited inimmunoprecipitation assays and SDS-PAGE analyses whenantisense oligonucleotides were used, in contrast to the find-ings with sense oligonucleotides (Fig. 4A). Furthemore, thebinding of AP-1 to its cognate DNA enhacer element wasevaluated using gel-shift analysis with a radiolabeled AP-1consensus sequence (Table 1).

DNA binding activity was examined in PMA-treated andnonstimulated cells. Treatment with PMA for 24 hours in-creased the intensity of the highest band (Fig. 4B), but didnot alter the intensity of the lowest band. The specificity ofthese bands was analyzed using a 100-fold excess of unla-beled wild-type oligonucleotide (Fig. 4B). The formation ofcomplexes was abrogated in the presence of wild-type butnot the mutant oligonucleotides (not shown). AP-1 bindingactivity was also modified by the polyclonal antisera spe-cific for c-fos and jun proteins (Fig. 4B). Preincubation withanti-jun antisera totally abrogated the uppermost band. Incontrast, incubation with anti-c-fos antisera partially re-duced the complex in the band shift assay. The lower bandwas not affected by any treatment.

In antisense experiments, U937 cells were previously ex-posed to 10 mM antisense or sense oligonucleotides for 1–4

Figure 4. (A) Effect of antisense c-fos oligonucleotides on fos proteininduction. Immunoprecipitation of fos protein was performed on lysatesfrom cells that were treated with sense or antisense oligonucleotides andthen stimulated with PMA, as described in the Materials and methods sec-tion. A faint band at 55 Kd can be detected in the antisense line (ANTIS) incomparison with the sense line (SENSE). This experiment was repeatedthree times. (B) Alterations in AP-1 binding activity in U937 cells incu-bated with antisense-c-fos or antisense-c-jun phosphorothioate oligonucle-otides. Radiolabeled oligonucleotide AP-1 was incubated with nuclearextract (3–5 mg) prepared from U937 cells (Basal lanes), PMA-treatedU937 cells in the absence of unlabeled specific competitor DNA (PMAlane) or in the presence of a 100-fold excess of unlabeled specific competi-tor (Comp lanes), U937 cells incubated with sense c-fos or sense c-junphosphorothioate oligonucleotide and then with PMA (Sense lanes), andU937 cells incubated with antisense c-fos or c-jun phosphorothioate oligo-nucleotide and then PMA (Antisense lanes). Supershifts were performedwith the radiolabeled probe, which was added to nuclear extract (5 mg)from PMA-treated U937 cells preincubated for 30 minutes at 48C with 0.5mL rabbit polyclonal antisera against c-fos (lane Ab c-fos), or jun-family(lane Ab jun). Each nuclear extract was also incubated with radiolabeledoligonucleotide oct-1 (Table 1), to verify that the amount nuclear extractused in each reaction was the same. Results from one of three experiments.

A. García et al./Experimental Hematology 27 (1999) 353–364 359

hours at 378C. Cells were transfected using either electropo-ration or liposome treatment and were then treated withPMA for 24 hours. As shown in Figure 4B, AP-1 activityincreased after PMA induction in cells transfected withsense oligonucleotides. Antisense c-fos and c-jun oligonu-cleotides partially affected AP-1 activity. In both cases,U937 cell differentiation and the pattern of expresson ofCD11/CD18 were identical to those in PMA-treated cells(Figs. 1 and 2A). These results suggest that during anti-c-fosor anti-c-jun treatment, residual AP-1 activity was alwaysdetected as a result of the complexity of the protein com-plexes that determine AP-1 binding activity. Furthemore,low binding activity was seen in some experiments in non-stimulated cells, which could not be removed in antisenseexperiments. For this reason we used double-stranded phos-phorothioate oligonucleotides containing the site for AP-1binding to eliminate residual AP-1 activity (Table 1, probeAP-1-ppt). In EMSA assays, the wild-type AP-1 sequencebound the AP-1 complex, whereas a mutated sequence didnot (Fig. 5A). With 7.5 mM of either wild-type AP-1 or mu-tated AP-1, U937 cells differentiated as seen in controls, and

the characteristic changes in morphologic and b2-integrinexpression occurred (Figs. 1 and 5B)

We also tested curcumin as a specific inhibitor of AP-1activity. This treatment significantly reduced AP-1 activityin EMSA assays (Fig. 3B, Curc2 lane). The results obtainedwere similar to those found with the decoy or antisensestrategies, and in contrast to the effect of PKC inhibitors,U937 cell differentiation and increment of b2-integrin ex-pression were not blocked (Fig. 3A).

U937 differentiation is specifically inhibited after sequestering of transcriptional PU.1 and NF-kB factorsWe tested the induction of b2-integrin molecules on the cellsurface of U937 cells after treatment with of NF-kB or PU.1double strand phosphorothioate oligodeoxynucleotides (Ta-ble 1). The specificity of the oligonucleotide for NF-kB wasassessed by mobility-shift assays using wild-type or mu-tated sequences (Fig. 6A). For PU.1, we used the sequencedescribed previously [34] and the specificity of the oligonu-cleotide for PU.1 was assased by mobility-shift assays (Fig.7A). We transfected or exposed U937 cells to 7.5 mM oligo-

Figure 5. (A) Characterization of AP-1 binding activity to the AP-1 double-stranded phosphorothioate oligonucleotide in PMA-treated U937 cells. Nuclearextract from PMA- treated U937 cells was incubated with the radiolabeled phosphorothioate oligonucleotide AP1-ppt (Table 1) in the absence of unlabeledspecific competitor DNA (PMA lane) or in the presence of a 100-fold excess of unlabeled specific competitor (COMP lane), in the presence of a 100-foldexcess of unlabeled nonspecific competitor AP-1-ppt-mut (MUT lane). Nuclear extract from PMA-treated U937 cells was also incubated with the radiola-beled phosphorothioate oligonucleotide AP-1-ppt mut (MUT* lane). (B) Expression of cell surface differentiation antigens in U937 cells incubated with theAP-1 double-stranded phosphorothioate oligonucleotide and PMA. U937 cells were transfected with 7.5 mM of either AP-1-ppt double-stranded phospho-rothioate oligonucleotide or Ap-1-ppt-mut double-stranded oligonucleotide. After 4 hours, cells were induced with PMA (20 ng/mL). Every 24 hours the cul-ture medium was replaced and a new dose of oligonucleotide was added; 72 hours after PMA induction, the cells were washed, stained and analyzed on aFACSort cytometer.

360 A. García et al./Experimental Hematology 27 (1999) 353–364

nucleotides for 1–2 hours at 378C and then cultured the cellsfor 72 hours with PMA. Exposure to the oligonucleotidewas not toxic and did not affect cell viability. The wild-typeNF-kB (NF-kB-ppt) clearly reduced CD11b expression tobelow the levels in nonstimulated cells, and led to moremoderate reductions in CD11a,c, CD18, and CD14 expres-sion. In contrast, the down-modulation of CD15 expressionwas unaffected (Fig. 6B). The mutated sequence had no ef-fect on b2-integrin expression (Fig. 6B). Cell attachmentand homotypic adhesion were also apparently blocked byNF-kB oligodeoxynucleotide treatment (Fig. 1).

The results obtained with a specific sequence corre-sponding to the PU.1 site of the CD11b promoter (PU.1-ppt,Table 1) differed from those with NF-kB oligonucleotide.Whereas adhesion to plastic and homotypic adhesion wasapparently blocked by both treatments (Fig. 1), in PU.1-treated cells the expression of CD11b and CD11c was mark-edly inhibited. Moderate inhibition was seen for CD18,CD14, while CD11a expression was weakly inhibited. Thelevels of expression of CD33 and CD15 were similar tothose in cells treated with the mutated sequence (Fig. 7B).

Our results also reveal a synergism between the tran-scription factors because the combination of either NF-kB/AP-1 and PU.1/AP-1 oligonucleotides reduced more signif-icantly than either alone (Fig. 8).

DiscussionThe exact role of the sequences in the promoter of CD11a,b, c, and CD18 involved in coordinated integrin inductionand stage-specific maturation of the myeloid cell line U937has not been fully resolved. The promoter regions for b2[15] and aM [22–24,37,38], aL [21,39], and aX [18] con-tain common putative regulatory sequences. The aim of ourstudy was to investigate the role of several transcriptionalfactors in PMA-induced differentiation in U937 cells. Pro-tein kinase C (PKC) is activated after PMA treatment and isbelieved to modulate the activity of transcription factorsthat bind to these cis elements. The PKC inhibitors stauro-sporine and bisindolylmaleimide significantly reduce AP-1and NF-kB induction, and completely block U937 cell dif-ferentiation induced by PMA. (Figs. 1 and 3A and 3B). Pre-

Figure 6. (A) NF-kB binding activity in PMA-U937 cells. Electrophoretic mobility shift assays (EMSA) showing the protein-DNA complexes that form theNF-kB double-stranded phosphorothioate oligonucleotide. Nuclear extract from PMA-treated U937 cells was incubated with the radiolabeled phosphorothio-ate oligonucleotide NF-kB-ppt (Table 1) in the absence of unlabeled specific competitor DNA (PMA lane), in the presence of a 100-fold excess of unlabeledspecific competitor (COMP lanes),or in the presence of a 100-fold excess of unlabeled nonspecific competitor NF-kB-ppt-mut (MUT lane). Nuclear extractfrom PMA-treated U937 cells was also incubated with the radiolabeled phosphorothioate oligonucleotide NF-kB-ppt mut (MUT* lane). (B) Variation in theexpression of cell surface differentiation antigens in U937 cells incubated with the NF-kB double-stranded phosphorothioate oligonucleotide and PMA. U937cells were transfected with 7.5 mM of either NF-kB-ppt double-stranded phosphorothioate oligonucleotide or NF-kB-ppt-mut double-stranded oligonucle-otide. After 4 hours, cells were induced with PMA (20 ng/mL). Every 24 hours the culture medium was changed and a new dose of oligonucleotide wasadded; 72 hours after PMA induction, the cells were washed, stained and analyzed on a FACSort cytometer.

A. García et al./Experimental Hematology 27 (1999) 353–364 361

vious studies have suggested that c-fos induction by phorbolesters is not necessary for HL60 or U937 differentiation[40,41]. In contrast, other studies have implicated AP-1 ac-tivity as a positive regulator of myeloid differentiation [42]and as a factor in the up-regulation of integrin expression[25]. Our results show that AP-1 activity is not sufficient toaccount for PMA induced b2-integrin transcriptional acti-vation. The induction of c-fos and c-jun, and the increase inAP-1 binding activity are not necessary or obligatory forU937 cell differentiation (Fig. 1). Our results with antisensestrategies showed that integrin activation and homotypiccell adhesion were not blocked. The reduction of c-fos pro-tein (Fig. 4A), and in c-fos- and c-jun-containing complexesin mobility shift assays during U937 cell differentiation wassignificant (Fig. 4B). However, binding was partially inhib-ited, probably because other proteins of the fos and jun fam-ily contribute to the generation of AP-1 binding activityduring diferentiation [25,43,44]. In fact, the patterns of inhi-bition were similar when the DNA binding reactions weresupplemented with antisera to fos protein (Fig. 4B). Resid-

ual binding activity was always seen, probably because thereduction in c-fos-containing complexes during U937 celldifferentiation was compensated by jun/jun dimers that alsorecognize the AP-1 element in the CD11c promoter [25].This residual binding activity may contribute to the absenceof inhibition of cell differentiation in antisense experiments.

With decoy strategies it is possible to sequester AP-1 andto inhibit the expression of genes regulated by this tran-scription factor [45,46]. However, our data show that stimu-lation of b2-integrin expression by PMA was unaffected incells exposed to phosphorothioate oligonucleotides contain-ing the AP-1 site (Fig. 5B). Similar results were obtainedwith curcumin, a specific AP-1 inhibitor (Fig. 3A). The pro-moter of CD11 and CD18 genes share similarities in organi-zation with the stromelysin gene and, in this case, the AP-1site is not necessary for the PMA-response [47,48]. Ourfindings are consistent with other studies that showed thatc-fos expression was not obligatory for U937 maturation,and that although the AP-1 binding site is involved in theactivity of integrin promoters, macrophage differentiation in

Figure 7. (A) PU.1 binding activity in PMA-U937 cells. Electrophoretic mobility shift assays (EMSA) showing the protein-DNA complexes that form thePU.1 double-stranded phosphorothioate oligonucleotide. Nuclear extract from PMA-treated U937 cells was incubated with the radiolabeled phosphorothioateoligonucleotide PU.1-ppt (Table 1) in the absence of unlabeled specific competitor DNA (PMA lane), in the presence of a 100-fold excess of unlabeled spe-cific competitor (COMP lanes), or in the presence of a 100-fold excess of unlabeled nonspecific competitor PU.1-ppt-mut (MUT lane). Nuclear extract fromPMA-treated U937 cells was also incubated with the radiolabeled phosphorothioate oligonucleotide PU.1-ppt mut (MUT* lane). (B) Expression of differenti-ation antigens in U937 cells incubated with the PU.1 double-stranded phosphorothioate oligonucleotide and PMA. U937 cells were transfected with 7.5 mMof either PU.1-ppt double-stranded phosphorothioate oligonucleotide or PU.1-ppt-mut double-stranded oligonucleotide. After 4 hours, cells were inducedwith PMA (20 ng/mL). Every 24 hours the culture medium was replaced and a new dose of oligonucleotide was added; 72 hours after PMA induction, thecells were washed, stained and analyzed.

362 A. García et al./Experimental Hematology 27 (1999) 353–364

vivo normally occurs in the absence of the AP-1 activity[41,49–51].

Phorbol esters also induce both the translocation ofNF-kB to the nucleus and the post-translational activationof PU.1 [26,52]. PU.1 elements are a common feature of thepromoters of the CD11 and CD18 genes. The critical rolefor PU.1 in mediating gene expression is shown by the abil-ity of double-stranded oligonucleotides containing the PU.1binding site to inhibit specific gene expression [34]. How-ever, contradictory results have been observed regarding therole of PU.1 binding during PMA-induced differentiation ofthe U937 cell line [26,53,54]. Our results suggest that PU.1contributes significantly to differentiation in U937. Thewild-type PU.1-binding phosphorothioate oligonucleotide,when added before PMA treatment, blocked adhesion prop-erties (Fig. 1) and clearly reduced integrin expression inU937 cells, in contrast with the mutated oligonucleotide(Fig. 7B). PU.1 is a positive regulator of the CD18 andCD11b gene promoters, and disruption of the PU.1 bindingsites significantly reduces promoter activity [55,56].

However, PU.1 appears to be a negative regulator of theCD11c promoter [57]. Our results also show a significantreduction in PMA-induced CD11c cell surface expression(Fig. 7B). In this context, the impaired expression of theCD18 gene could affect cell surface expression of the CD11subunits, in a manner similar to that in inherited Leu-CAMdeficiency [17].

Integrin expression was also inhibited by competitivebinding of NF-kB factors. However, putative binding se-quences have been identified only in the CD11a promoter,but have not been found for CD11b. Interestingly, the ex-

pression of CD11b fell to below basal levels (Fig. 6B). Inagreement with our findings, antisense oligonucleotides tothe p65 subunit of NF-kB block CD11b expression but donot inhibit gene expression [58]. We also found a strong in-hibition in the expression of CD11a, CD11c, CD18, CD33,and CD14. It has been suggested that p65 NF-kB oligonu-cleotides may be blocking transcription of some factor(s)important for the release of intracellular granules containingpreformed CD11b or, alternatively, p65 may be affecting therelease of CD11b from granules by nontranscriptionalmechanisms [58].

We suggest that NF-kB activation is a necessary eventthat affects the entire differentiation program in U937. Al-though both PU.1 and NF-kB oligonucleotides clearly af-fected cell surface adhesion, they appear to be responsiblefor only part of the induction of CD11/CD18 expressionduring U937 cell differentiation (Figs. 6B and 7B). Theseresults may be explained by insufficient competitive bind-ing of the nuclear factors under our experimental condi-tions. Alternatively, it is also possible that the developmen-tal regulation of integrin expression is not fully under thecontrol of a nuclear factor alone, and treatment with a singleoligonucleotide is not sufficient to induce complete inhibi-tion. Thus, the absence of nuclear factor may be compen-sated by the presence of other positive transacting factors[56]. For this reason, combinations of either NF-kB/AP-1 orPu.1/AP-1 oligonucleotides in cotransfection experimentswere also done. The combinations of NF-kB and AP-1 pro-duced synergistic inhibition (Fig. 8). Both transcriptionalfactors target unique enhancer elements; however, its hasbeen demonstrated that the cross-coupling of these tran-

Figure 8. Sinergistic effects of decoy oligonucleotides on PMA-induced U937 cells. The combined use of either AP-1/NF-kB or AP-1 /PU.1 oligonucle-otides evidenced a stronger reduction in integrin cell surface expression in comparison with the effect obtained with each oligonucleotide separately. j Basal,—— Mut1PMA, —— NF-kB1PMA, . . . . NF-kB1AP11PMA (left); j Basal, —— Mut1PMA, —— PU.11PMA, . . . . PU.11AP11PMA (right)

A. García et al./Experimental Hematology 27 (1999) 353–364 363

scription factors pontentiates biologic functions in vivothrough either kB or AP-1 enhancer elements [59]. In addi-tion, Ets transcription factors also cooperate with AP-1 fortranscriptional activation [60–62].

The development of stable oligonucleotide competitorsmakes it possible to study individual nuclear transcriptionalfactors by blocking the expression of genes involved inpathologic conditions. In this context, the presence of multi-ple regulatory elements in the gene promoter makes it nec-essary to obtain a detailed understanding of the real implica-tion of nuclear factors in gene expression or repression.Therefore, reagents that target integrin expression can beused to study cellular differentiation and function, and mayhave potentially important therapeutic applications.

AcknowledgmentsWe thank Dr. R. Bravo for the kind gift of the anti-jun and anti-fosantisera. We also thank Ms. Carmen Amezcua for technical assis-tance, and Karen Shashok for improving the English style of themanuscript. Pilar Jiménez and Angel García were supported byB.A.E. fellowships 97/5049 and 97/5281. This research was sup-ported by the Fondo de Investigaciones Científicas, project FIS 96/1497; and Servicio Andaluz de Salud, project SAS 218/97.

References1. Rovera G, O’Brian TA, Diamond L (1979) Induction of dferentiation

in human promyelocytic leukemic cells by tumor promoters. Science204:868

2. Lubert M, Herrmann F, Koeffler HP (1991) Expression and regulationof myeloid-specific genes in normal and leukemic myeloid cells.Blood 77:909

3. Larsson LG, Ivhed I, Gidlund M, Pettersson Ulf, Vennstrom B, Nils-son K (1988) Phorbol ester-induced terminal differentiation is inhib-ited in human U-937 monoblastic cells expressing a v-myc oncogene.Proc Natl Acad Sci U S A 85:2638

4. Hass R, Bartels H, Topley N, Hadam M, Köhler L,Goppelt-Strübe M,Resch K (1989) TPA-induced differentiation and adhesion of U937cells: changes in ultrastructure, cytoskeletal organization and expres-sion of cell surface antigens. Eur J Cell Biol 48:282

5. Nishizuka Y (1984) The role of protein kinase C in cell surface signaltransduction and tumour promotion. Nature 308:693

6. Pedrinaci S, Ruiz-Cabello F, Gomez O, Collado A, Garrido F (1990)Protein kinase C-mediated regulation of the expression of CD14 andCD11/CD18 in U937 cells. Int J Cancer 45:294

7. Ghosh S, Baltimore D (1990) Activation in vitro of NF-kappa B byphosphorylation of its inhibitor I kappa B. Nature 344:678

8. Angel P, Imagawa M, Chiu R, Stein B, Imbra RJ, Rahmsdorf HJ, JonatC, Herrlich P, Karin M (1987) Phorbol ester-inducible genes contain acommon cis element recognized by a TPA-modulated trans-acting fac-tor. Cell 49:729

9. Baeuerle PA, Baltimore D (1988) I kappa B: a specific inhibitor of theNF-kappa B transcription factor. Science 242:540

10. Chiu R, Imagawa M, Imbra RJ, Bockoven JR, Karin M (1987) Multi-ple cis- and trans-acting elements mediate the transcriptional responseto phorbol esters. Nature 329:648

11. Bos TJ, Bohmann D, Tsuchie H, Tjian R, Vogt PK (1988) v-jun en-codes a nuclear protein with enhancer binding properties of AP-1. Cell52:705

12. Williams T, Admon A, Luscher B, Tjian R (1988) Cloning and expres-sion of AP-2, a cell-type-specific transcription factor that activates in-

ducible enhancer elements. Genes Dev 2:155713. Imagawa M, Chiu R, Karin M (1987) Transcription factor AP-2 medi-

ates induction by two different signal-transduction pathways: proteinkinase C and cAMP. Cell 51:251

14. Mercurio F, Karin M (1989) Transcription factors AP-3 and AP-2 in-teract with the SV40 enhancer in a mutually exclusive manner. EMBOJ 8:1455

15. Bottinger EP, Shelley CS, Farohkzad OC, Arnaout MA (1994) The hu-man b2 integrin CD18 promoter consists of two inverted Ets cis ele-ments. Mol Cell Biol 14:2604

16. Bellon T, López-Rodriguez C, Rubio MA, Jochems G, Bernabeu C,Corbí AL (1994) Regulated expression of p150,95 (CD11c/CD18;X/2) and VLA-4 (CD49d/CD29; 4/ 1) integrins during myeloid celldifferentiation. Eur J Immunol 24:41

17. Arnaout MA (1990) Structure and function of the leukocyte adhesionmolecules CD11/CD18. Blood 75:1037

18. López-Cabrera M, Nueda A, Vara A, García-Aguilar A, Tugores A,Corbí AL (1993) Characterization of the p150,95 leukocyte integrin asubunit (CD11c) gene promoter. Identificacion of cis-acting se-quences. J Biol Chem 268:1187

19. Nueda A, López-Cabrera A, Vara A, Corbí AL (1993) Characteriza-tion of the CD11a ( L, LFA-1 ) integrin gene promoter. J Biol Chem268:19305

20. Shelley CS, Farohkzad OC, Arnaout MA (1993) Identification of acell-specific and developmentally regulated nuclear factor that directmyeloid and lymphoid expression of CD11a gene. Proc Natl Acad SciU S A 90:5364

21. Cornwell RD, Gollahon KA, Hickstein DD (1993) Description of theleukocyte function-associated antigen 1 (LFA-1 or CD11a) promoter.Proc Natl Acad Sci U S A 90:4221

22. Pahl HL, Rosmarin AG, Tenen DG (1992) Characterization of the my-eloid-specific CD11b promoter. Blood 79:865

23. Shelley CS, Arnaout MA (1991) The promoter of CD11b gene directsmyeloid-specific and developmentally regulated expression. Proc NatlAcad Sci U S A 88:10525

24. Hickstein DD, Baker DM, Gollahon KA, Back AL (1992) Identifica-tion of the promoter of the myelomonocytic leukocyte integrin CD11b.Proc Natl Acad Sci U S A 89:2105

25. López-Rodríguez C, Kluin-Nelemana HC, Corbí AL (1996) AP-1 reg-ulates the basal and developmentally induced transcription of theCD11c leukocyte integrin gene. J Immunol 156:3780

26. Carey J, Posekany JK, Devente JE, Pettit GR, Ways Kirk (1996) Phor-bol esters-stimulated phosphorylation of PU.1: association with leuke-mic cell growth inhibition. Blood 87:4316

27. David P, Elliot L, Harris W, Hill C, Hurst S, Keech E, Kumar M, Law-ton G, Nixon J, Wilkinson MS (1992) Inhibitors of protein kinase C.2.substituted bisindolymaleimides with improved potency and selectiv-ity. J Med Chem 35:944

28. Huang TZ, Lee SC, Lin JK (1991) Supression of c-jun/AP-1 activationby an inhibitor of tumor promotion in mouse fibroblast cells. Proc NatlAcad Sci U S A 88:5292

29. Bradshaw TD, Gescher A, Pettit GR (1992) Modulation by stauro-sporin of phorbol-ester-induced effects on growth and protei kinase Clocalization in A549 human lung carcinoma cells. Int J Cancer 51:144

30. Hanazawa S, Takeshita A, Amano S, Semba T, Nirazuka T, Katoh H,Kitano S (1993) Tumor necrosis factor-alpha induces expression ofmonocyte chemoattractant JE via fos and jun genes in clonal osteo-blastic MC3T3-E1 cells. J Biol Chem 268:9526

31. Iyer RP, Egan W, Regan JB, Beaucage SL (1990) 3 H-1,2-Benzodithi-ole-3-one 1,1-dioxide as an improved sulfurizing reagent in the solid-phase synthesis of oligodeoxyribonucleoside phosphorothioates. J AmChem Soc 112:1253

32. Schereiber E, Mathias P, Muller MM, Schaffner W (1989) Rapid de-tection of octamer binding proteins with “mini-extract”, prepared froma small number of cells. Nucleic Acids Res 17:6419

33. Abril E, Mendez RE, García A, Serrano A, Cabrera T, Garrido F,

364 A. García et al./Experimental Hematology 27 (1999) 353–364

Ruiz-Cabello F (1995) Characterization of a gastric tumor cell line de-fective in MHC class I inducibility by both a- and g-interferon. TissAntigens 47:391

34. Voso MT, Burn TC, Wulf G LIM B Leone G, Tenen DG (1994) Inhi-bition of hematopoiesis by competitive binding of transcription factorPU.1. Proc Natl Acad Sci U S A 91:7932

35. Mitchell RL, Zokas L, Schreiber RD, Verma IM (1985) Rapid induc-tion of the expression of proto-oncogene fos during human monocyticdifferentiation. Cell 40:209

36. Lamph WW, Wamsley P, Sassone-Corsi P, Verma IM (1988) Induc-tion of proto-oncogene JUN AP-1 by serum and TPA. Nature 334:629

37. Rosmarin AG, Weil SC, Rosner GL, Griffin JD, Arnaout MA, TenenDG (1989) Differential expression of CD11b/CD18 (Mac1) and my-eloperoxidase genes during myeloid differentiation. Blood 73:131

38. Pahl HL, Rosmarin AG, Tenen DG (1992) Characterization of the my-eloid-specific CD11b promoter. Blood 79:865

39. Back AL, Gollahon KA, Hickstein DD (1992) Regulation of expres-sion of the leukocyte integrin CD11a (LFA-1) molecule during differ-entiation of HL-60 cells along the monocyte/macrophage pathway. JImmunol 148:710

40. Calabretta B (1987) Dissociation of c-fos induction from macrophage dif-ferentiation in human myeloid leukemic cell lines. Mol Cell Biol 7:769

41. Mitchell RL, Henning-Chubb C, Huberman E, Verma IM (1986) c-fosExpressin is neither sufficient nor obligatory for differentiation ofmonomyelocytes to macrophages. Cell 45:497

42. Lord KA, Abdollahi A, Hoffman-Liebermann B, Liebermann DA(1993) Proto-oncogens of the fos/jun family of transcription factors arepositive regulators of myeloid differentiation. Mol Cell Biol 13(2):841

43. Kovary K, Bravo R (1991) Expression of different jun and fos proteinsduring the G0 to G1 transition in mouse fibroblasts: in vitro and in vivoassociations. Mol Cell Biol 12:5015

44. Kovary K, Bravo R (1992) Existence of different fos/jun complexesduring transition and during exponential growth in mouse fibroblasts.Mol Cell Biol 12:5015

45. Bielinska A, Shivdasani RA, Zhang LO, Nabel GJ (1990) Regulationof gene expression with double-stranded phosphorothioate oligonucle-otides. Science 250:997

46. Eck SL, Perkins ND, Carr DP, Nabel GJ (1993) Inhibition of phorbolester-induced cellular adhesion by competitive binding of NF-kB invivo. Mol Cell Biol 13:6530

47. Wasylyk C, Gutman A, Nicholson R, Wasylyk B (1991) The c-Ets on-coprotein activates the stromelysin promoter through the same ele-ments as several non-nuclear oncoproteins. EMBO J 10:1127

48. Buttice G, Quinones S, Kurkinen M (1991) The AP-1 site is requiredfor basal expression but is not necessary for TPA-response of the hu-

man stromelysin gene. Nucleic Acids Res 19:372349. Grigoriadis AE, Wang Z, Cechini MG, Hofstetter W, Felix R, Fleish

HA, Warner EF (1994) c-Fos: a key regulator of osteoclast-macro-phage lineage determination and bone remodeling. Science 266:443

50. Wang Z, Liang J, Schellander K, Wagner EF, Grigoriadis AE (1995)C-fos-induced osteosarcoma formation in transgenic mice: cooperativ-ity with c-jun and the role of endogenous c-fos. Cancer Res 55:6244

51. Shivdasani RA, Orkin SH (1996) The transcriptional control in he-matopoiesis. Blood 87(10):4025

52. Baeuerle PA (1991) The inducible transcription activator NF-kB: reg-ulation by distinct protein subunits. Biochim Biophys Acta 1072:63

53. Chen HM, Pahl HL, Scheibe RJ, Zhang DE, Tenen DG (1993) TheSp1 transcription factor binds the CD11b promoter specifically in my-eloid cells in vivo and is essential for myeloid-specific promoter activ-ity. J Biol Chem 268:8230

54. Pahl HL, Scheibe RJ, Zhang DE, Chen HM, Galson DL, Maki RA,Tenen DG (1993) The proto-oncogene PU.1 regulates expression ofthe myeloid-specific CD11b promoter. J Biol Chem 268:5014

55. Rosmarin AG, Caprio D, Levy R, Simkevich C (1995) CD18 (beta 2leukocyte integrin) promoter requires PU.1 transcription factor for my-eloid activity. Proc Natl Acad Sci U S A 92:801

56. Rosmarin AG, Caprio D, Kirsch DG, Handa H, Simkevich C (1995)GABP and PU.1 compete for binding, yet cooperate to increaseCD18(beta 2 leukocyte integrin) transcription. J Biol Chem 270:23627

57. López-Rodriguez C, Corbí AL (1997) PU.1 negatively regulates theCD11c integrin gene promoter through recognition of the major tran-scriptional start site. Eur J Immunol 27:1843

58. Sokoloski JA, Sartorelli AC, Rosen CA, Narayanan R (1993) Anti-sense oligonucleotides to the p65 subunit of NF-kB block CD11b ex-pression and alter adhesion properties of differentiated HL-60 granulo-cytes. Blood 82(2):625

59. Stein B, Baldwin Jr AS, Wallard DW, Greene WC, Angel P, HerrlichP (1993) Cross-coupling of the NF-kB p65 and Fos/Jun transcriptionalfactors produces potentiated biological function. EMBO J 12(10):3879

60. Wasylyk B, Wasylyk C, Flores P, Begue A, Leprince D, Stehenli D(1990) The c-ets proto-oncogenes encode transcriptional factors thatcooperates with c-fos and c-jun for transcriptional activation. Nature346:191

61. Wu H, Moulton K, Horval A, Parik S, Glass CK (1994) Combinatorialinteractions between AP-1 and Ets domain protein contribute to the de-velopmental regulation of the macrophage scavenger receptor gene.Mol Cell Biol 14:2129

62. Bassuk AG, Leiden JM (1995) A direct physical association betweenETS and AP-1 transcription factors in normal human T cells. Immu-nity 3:223